Robert D. Treitman

Personal exposures to respirable particulates and implications for air pollution epidemiology.

Measurements of personal exposures to respirable particles (RSP) were obtained from nonsmoking adults living in two rural Tennessee communities. Personal exposure measurements were compared to simultaneously collected indoor (home) and outdoor concentrations. Personal exposures were higher than, had a greater variance than, and were uncorrelated with outdoor concentrations. Household smoking was found to be a substantial contributor to personal RSP exposure. Regressions of indoor concentrations on paired personal exposures explained 11-87% of the variance in exposure depending on employment subgroup and household smoking. A deterministic, predictive model based on the time spent in four microenvironments and measured concentrations explained 64% of the variance in personal exposure. Ambient concentrations provide poor prediction of personal exposure to undifferentiated respirable size particles. Air pollution epidemiological investigations must consider the importance of indoor environments in estimating subject exposures. Further, the chemical/elemental compositions of indoor concentrations and personal exposures are likely to be different from ambient concentrations. This study indicates the potential for misclassification and misassociation of exposures that are likely to result in relying upon ambient, community-based particle measurements.

Air contaminants encountered by firefighters

The concentrations of eight air contaminants suspected of causing acute and chronic health problems for firefighters were measured in over 200 fires in the City of Boston using a personal air sampler. Threatening concentrations of both carbon monoxide and acrolein were found in a small proportion of the fires. Less hazardous levels of hydrogen chloride, hydrogen cyanide, nitrogen dioxide and carbon dioxide were also noted. Benzene was found in most fires, but at concentrations well below those expected to cause acute injury. The air sampling data have application in treatment of smoke inhalation victims, development of firefighting strategies and selection of respiratory protection devices.

Exposures to Respirable, AirbornePenicillium from a Contaminated Ventilation System: Clinical, Environmental and Epidemiological Aspects

Symptoms compatible with hypersensitivity pneumonitis (HP) in two of fourteen employees in a clerical office prompted an investigation of their work environment. Forced-air heater-cooler units which had not been properly maintained were implicated when they were found to be grossly contaminated with predominantlyPenicillium molds. Air-sampling for viable, respirable-size particulates in the affected office and an unaffected office in the same building demonstrated a 5O- to 80-fold excess in the number of colony-forming-units per cubic meter of air in the affected office. Persistent alveolitis was documented by repeated bronchoalveolar lavage, gallium scan, and other studies in one affected worker whose peripheral lymphocytes underwent blast transformation in response toPenicillium antigens obtained by air-sampling in the work environment. The other affected worker had asthma, presumably exacerbated by exposures to a variety of inhaled environmental irritants and antigens. Despite a documented reduction of...

Personal exposure to respirable particles: a case study in Waterbury, Vermont

Abstract A study to assess personal exposure to respirable particles was conducted during January to March 1982 in Waterbury, Vermont. Forty-eight nonsmoking volunteers carried Harvard/EPRI personal samplers every other day for two weeks. Simultaneous measurements with similar monitors were made inside and outside each participants home. Findings indicate that outdoor (ambient) particle levels were not an important determinant of personal exposure, while in-home concentrations accounted for 25–30% of the variation in personal values. A linear regression technique was used to estimate respirable particle concentrations in three micro-environments where measurements were not available. These values were combined with data on time activities and observed outdoor and in-home concentrations to construct a simple time-weighted exposure model. Predicted exposure using this approach agreed well with measured values, however, the validity and suitability of estimated coefficients for applications to other communities and different times of year has not been established.

Effects of residential wood combustion on indoor air quality: a case study in Waterbury, Vermont

Abstract An indoor/outdoor monitoring study was conducted during January–March 1982 in Waterbury, Vermont. Respirable particle measurements were made inside and outside 24 homes (all occupants were nonsmokers), 19 with wood-burning appliances and 5 without. Data were also obtained on seasonal air-exchange rate, heating fuel consumption, and relevant home characteristics. Findings indicate that indoor particle levels are consistently higher than outdoor values regardless of heating-fuel type. No statistical difference was observed between 24-h average respirable particle levels in wood- and nonwood-burning homes. A linear regression model, incorporating information on air-exchange rate, house volume, fuel use, and outdoor levels, accounted for about 20% of the variance in indoor particle concentrations.

The Boston residential NO2 characterization study—II. Survey methodology and population concentration estimates

Abstract A recently completed year-long study of NO2 and air exchange rates in over 500 homes in the Boston Metropolitan area provides data to quantify the component of total NO2 exposures attributable to indoor sources, especially to gas-fired appliances. The approach of this work was to provide field data for validation or refinement of exposure models developed in previous, related work. For an indoor characterization field study, sample sizes of 450 gas- and 150 electric-range-equipped housing units were selected based on: (1) modeled estimation of precision and stability of parameter estimates using various sample sizes; (2) calculations including anticipated attrition from one monitoring period to the next. The sample was selected using standard area probability sampling to allow extrapolations of survey sample results to the larger population. The survey design included stratification by range fuel and area clustering for sampling and logistical efficiency. This paper presents the sampling results and field work progress through the year, with a discussion of response rates typical for exposure monitoring investigations. Monitoring results provide NO2 concentration data to evaluate the overall success of the survey implementation. A series of analyses isolate and quantify the standard errors of distribution estimates. Using sample data weighted for stratification, population exposure distribution parameter estimates are presented. Overall, analyses indicate that key model assumptions are valid. The relatively low standard errors of exposure parameters indicate that the design used in the study was relatively efficient. This illustrates the utility of standard survey research methodology in exposure assessment problems.

Winter air quality in a wood-burning community: A case study in Waterbury, Vermont

Abstract The recent upsurge in residential wood combustion has raised questions about potential adverse effects on ambient air quality and public health. Before policymakers can make informed and rational decisions about the need for government intervention, more information is needed concerning the nature and extent of the problem. This paper presents findings from the 1982 Harvard Wood-Burning Study in Waterbury, Vermont. Waterbury, a rural community of about 2000 people, was an ideal location for this investigation because: (1) half of the private residences are heated with wood fuel; (2) frequent winter temperature inversions promote pollution buildup in the valley; (3) there are no major industrial sources and (4) the Vermont Agency of Environmental Conservation has compiled a detailed wood-burning inventory. The ambient air monitoring study, from January to March 1982, emphasized measurements of total, inhalable and respirable particulate matter. Results indicate that 60–70% of the Waterbury aerosol was composed of particles less than 2.5 μm. A combination of indirect evidence suggests that wood burning was the major source of airborne particles in residential sections of the town. Dramatic diurnal variations in particulate concentrations were observed, with peak values at night exceeding afternoon levels by 5- to 10-fold. Both meteorology and emission patterns contributed to observed fluctuations.

Characterization of indoor air quality in wood-burning residences

Abstract Findings from a residential wood-burning study in Waterbury, VT, are presented, with emphasis on indoor-outdoor comparisons of particle mass, size distribution, and composition. The air monitoring program was carried out from January to March 1982 to evaluate the impact of residential wood combustion on indoor and outdoor air quality. Indoor and outdoor data on respirable particle concentrations were obtained for 24 residences, 19 with wood-burning appliances. This paper focuses on 6 of these homes which were selected for more intensive study, including matched indoor-outdoor particle measurements to determine elemental composition, individual polycyclic aromatic hydrocarbon concentrations, and organic- and elemental-carbon content. Data are presented comparing particle-phase organic and elemental concentrations between indoor and outdoor environments.